Metabolomic and functional profiling of Lactiplantibacillus plantarum strains reveals distinct probiotic, immunomodulatory, and antimicrobial signatures

Abstract

Abstract

Lactiplantibacillus plantarum strains are increasingly recognized for their combined probiotic and antimicrobial activities, offering potential applications in gut health management and pathogen control. This study characterized the intracellular (Met-Int) and extracellular (Met-Ext) metabolomic profiles of L. plantarum UTNGt2 (Gt2), UTNGt3 (Gt3), and UTNGt28L (Gt28L) isolated from tropical fruits and evaluated their probiotic, antimicrobial, cytotoxic, and immunomodulatory properties in vitro. Metabolomic profiling was performed using liquid chromatography–tandem mass spectrometry (LC–MS/MS) with a SWATH (Sequential Windowed Acquisition of All Theoretical Fragment Ion Mass Spectra) acquisition method. Cytotoxicity and cell viability were assessed by MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) and LDH (lactate dehydrogenase) release assays in colon epithelial cells, while cytokine responses (IL-10, IL-1β) were quantified to determine immunomodulatory effects. Antimicrobial mechanisms were examined by scanning and transmission electron microscopy (SEM/TEM) on Staphylococcus aureus ATCC1026. LC–MS/MS identified 117 Met-Int and 32 Met-Ext across the three strains, revealing shared metabolites (e.g., l-tryptophan, adenosine) and distinct strain-specific compounds (e.g., harmine, lincomycin, baicalein) associated with bioactivity. Pathway enrichment analysis indicated four enriched pathways in Gt2, eight in Gt3, and ten in Gt28L, reflecting differential specialization in amino-acid, carbohydrate, and cofactor metabolism. Gt3 exhibited the most diverse antimicrobial metabolite repertoire, whereas Gt28L showed the strongest anti-inflammatory effect, increasing IL-10 secretion by ~ 6.5-fold and reducing IL-1β by ~ 50% compared with control cells. All strains maintained > 85% cell viability with minimal LDH release. SEM/TEM analysis confirmed that Met-Ext fractions caused membrane disruption and intracellular damage in S. aureus. Overall, these results demonstrate strain-specific probiotic, antimicrobial, and immunomodulatory signatures, identifying Gt3 and Gt28L as promising candidates for therapeutic or food-grade applications.

Key points

• Distinct L. plantarum strains show unique metabolite profiles with probiotic potential.

• Gt3 and Gt28L strains exhibit strong antimicrobial and anti-inflammatory activity.

• Metabolite extracts disrupt pathogens while preserving epithelial cell viability.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00253-026-13706-3.

Introduction

Lactiplantibacillus plantarum is a highly versatile and extensively studied member of the lactic acid bacteria (LAB) group, widely recognized for its substantial probiotic potential (Echegaray et al. 2022). Its ability to colonize diverse ecological niches, including fermented foods and the gastrointestinal tracts of humans and animals, has positioned this species as a cornerstone in food biotechnology and health-related applications (Tang et al. 2023). The probiotic and immunomodulatory properties of L. plantarum are largely attributed to its broad metabolic repertoire, which includes organic acids such as lactic acid and short-chain fatty acids (SCFAs) that contribute to pathogen inhibition, gut homeostasis, and anti-inflammatory effects (Fidanza et al. 2021). In addition, many strains produce bacteriocins, particularly plantaricins, exhibiting broad-spectrum antimicrobial activity against bacteria, fungi, and certain viruses (Tang et al. 2023), as well as exopolysaccharides (EPS) that promote adhesion to intestinal epithelial cells, support biofilm formation, and modulate host immune responses (Fidanza et al. 2021; Tang et al. 2023). Other metabolites, including gamma-aminobutyric acid (GABA) and indole-3-lactic acid (ILA), further expand the functional activity of L. plantarum by contributing to stress regulation, gut barrier integrity, neurological support, and immune modulation (Guamán et al. 2024). Some strains are also capable of synthesizing vitamins and additional bioactive molecules that enhance host nutritional status (Behera et al. 2018). Importantly, the production and biological activity of these metabolites are strongly influenced by strain identity, environmental conditions, and nutrient availability (Bhattacharya et al. 2025). Variations in metabolite profiles and organic acid ratios are reflected in strain-dependent clinical effects (Werning et al. 2022; Yilmaz et al. 2022). For example, L. plantarum 299v has demonstrated benefits in improving gastrointestinal symptoms, oral health, and systemic inflammation (Chatsirisakul et al. 2025), whereas L. plantarum MF1298 has been reported to exacerbate symptoms of irritable bowel syndrome (Ligaarden et al. 2010). Other strains, including IS-10506, have shown immunomodulatory effects in atopic dermatitis (Prakoeswa et al. 2022), LB931 reduced pathogenic colonization in the vaginal environment (Rönnqvist et al. 2005), and CCFM8724 decreased early childhood caries (Zhang et al. 2020). Neurobehavioral effects have been described for L. plantarum PS128 (Liu et al., 2019), while strain OLL2712 has been reported to support gut barrier function (Watanabe-Yasuoka et al. 2023). Other strains, including Z22 and 19-2, have been associated with anti-inflammatory, antimicrobial, metabolic, and immunoregulatory activities (Cristofori et al. 2021; Wang et al. 2024). Together, these studies suggest that L. plantarum encompasses strains with diverse functional attributes that may be relevant for the development of targeted probiotic applications (Tavallaei et al. 2025).

Given that such functional traits are closely linked to metabolic capacity and ecological adaptation, the isolation source is likely to influence probiotic performance. In this regard, non-dairy and traditionally fermented foods, including fruits, vegetables, and spontaneous fermentations, represent rich reservoirs of LAB with substantial metabolic diversity and probiotic potential (Swain et al. 2014; Adesulu-Dahunsi et al. 2018). For example, Lactobacillus species isolated from cocoa (Theobroma cacao L.) fermentations have demonstrated protective activity against Salmonella typhimurium (Oliveira et al. 2018), underscoring the value of exploring non-conventional fermentation matrices. These observations support expanding probiotic discovery beyond traditional dairy systems to include bacteria adapted to spontaneous fermentations that reflect the native microbiota of their ecological niches. Accordingly, our previous work has focused on bacterial communities associated with wild Amazonian fruits from Ecuador as part of a broader effort to identify and develop novel probiotic candidates. From these substrates, we isolated and characterized several LAB strains exhibiting tolerance to physiologically relevant acidic and bile salt conditions, along with in vitro antimicrobial activity (Benavidez et al. 2016). Among these, strains Gt2, Gt3, and Gt28L emerged as particularly promising, combining robust stress tolerance with pronounced antimicrobial properties (Tenea & Ortega 2021). Strain Gt2 was isolated from white cocoa (Theobroma grandiflorum), while Gt3 and Gt28L originated from Chrysophyllum oliviforme and Chrysophyllum caimito, respectively. The distinct ecological backgrounds of these strains enable the assessment of strain-specific metabolic variability. Comprehensive profiling of intracellular and extracellular metabolite pools in these tropical L. plantarum strains represents a novel extension of previous omics-based studies (Echegaray et al. 2022) and enables deeper insight into how environmental origin shapes functional potential. Based on our earlier findings, the present study investigates the metabolite profiles of strains Gt2, Gt3, and Gt28L, with a focus on identifying bioactive compounds involved in gut health promotion and pathogen inhibition. By integrating metabolite profiling with functional analyses, we aim to elucidate the relationship between ecological origin, metabolic diversity, and probiotic potential, thereby identifying promising candidates for targeted applications in human and animal health.

Materials and methods

Bacterial strains

Bacterial strains were isolated from wild fruits of Theobroma grandiflorum, C. oliviforme, and C. caimito from the Cuyabeno rainforest (Sucumbíos Province, Ecuador) using the dilution plating method. Isolates were incubated on De Man, Rogosa, and Sharpe (MRS) agar (Difco, Detroit, MI, USA) at 37 °C for 72 h under anaerobic conditions. Candidate strains were screened for Gram reaction, motility, catalase and indole production, spore formation, gas production from glucose, and tolerance to acidic pH (2.5–4.0), bile (0.3–1%), and antibacterial activity against Gram-positive and Gram-negative bacteria. Lactiplantibacillus plantarum strains Gt2 (BioProject PRJNA705232; BioSample SAMN18053630), Gt3 (BioProject PRJNA1116628; BioSample SAMN48548443), and Gt28L (BioProject PRJNA1116628; BioSample SAMN49560224) demonstrated the strongest antibacterial activity and stress tolerance and were selected for further study. Strains were maintained on MRS agar and stored at −80 °C in 20% glycerol. They are preserved in the CCMBIOGEM culture collection (BIOGEM Laboratory, UTN, Ecuador) and are available for research upon request.

Metabolite extraction

For metabolite extraction, cultures were grown in MRS broth at 37 °C for 24 h, then centrifuged at 13,000 × g for 30 min at 4 °C. The cell-free supernatant (Met-Ext) was filtered through 0.22-µm membranes (#STF020025H, Chemlab Group, USA) and stored at 4 °C. Intracellular metabolites were obtained by resuspending the bacterial pellet in methanol: water (4:1, v/v; HPLC–MS grade), followed by three freeze-thaw cycles in liquid nitrogen, incubation at 35 °C for 5 min, vortexing, and sonication (20 kHz, 2 min on ice). After centrifugation (13,000 × g, 10 min, 4 °C), the supernatant (Met-Int) was collected. Finally, the extracts were lyophilized and stored for further analysis.

LC–MS/MS analysis, metabolites identification, and enrichment pathway prediction

Metabolite profiling was performed as previously described (Molina et al. 2025). Briefly, samples (1 mg/mL) were centrifuged at 17,000 × g for 15 min at 4 °C, and the clear supernatants were analyzed using an AB SCIEX Triple TOF 5600+ mass spectrometer (Sciex, Canada) coupled to a nanoACQUITY UPLC system (Waters, USA). Chromatographic separation was achieved on an Eksigent 5C18-CL-120 column (300 µm × 150 mm) with a linear gradient of 5–80% acetonitrile (0.1% formic acid) over 90 min at 5 µL/min and 55 °C. Data were acquired in both positive and negative electrospray ionization modes, under optimized source conditions. Untargeted analysis was carried out using SWATH-MS (60 variable windows), with MS1 scans (m/z 100–1250) and MS2 scans (m/z 100–2000) at accumulation times of 150 ms and 30 ms, respectively, with the ion source parameters as follows: GS1: 15, GS2: 0, CUR: 25, TEM: 0, and ISVF: 5500 V. The applied source parameters represent a low-temperature, low-nebulizer configuration tailored for micro-flow LC-ESI, where efficient desolvation is achieved without auxiliary heating or GS2 gas, thereby limiting in-source fragmentation and preserving thermolabile metabolites. Data were processed with MS-DIAL v5.3.240719, and metabolites were identified by matching MS/MS spectra against the MSP libraries (https://systemsomicslab.github.io/compms/msdial/main.html#MSP) (accessed on 15 April 2025). For metabolites’ identification, MS-DIAL was used with the following parameters: retention time range of 1–90 min; MS1 mass range of 100–1250 Da and MS2 mass range of 100–2000 Da; minimum peak width of 5 scans; minimum peak height of 1000 (amplitude); smoothing level set to 3 scans; MS2 spectrum cutoff at 10 (absolute amplitude); mass slice width of 0.05 Da; no exclusion mass list applied; retention time tolerance of 0.1 min; MS1 mass accuracy tolerance of 0.01 Da; and MS2 mass accuracy tolerance of 0.05 Da. A minimum matched spectrum similarity of 70% was required. Additionally, mass accuracy tolerances of 0.01 Da for MS1 and 0.025 Da for MS2 were applied. Metabolites identified via LC–MS were annotated and mapped to metabolic pathways using MetaboAnalyst Pathway Analysis 6.0, which utilizes databases such as HMDB, PubChem, and KEGG to convert compound identifiers and determine their biological roles, with pathway enrichment significance adjusted using false discovery rate (FDR) (Benjamini and Hochberg 1995) as a multiple testing correction (Pang et al. 2024; Kanehisa et al. 2024).

Cytotoxicity and viability assessment of metabolites on Caco-2 cells

Caco-2 (HTB-37™, ATCC) cells were seeded at 1 × 10⁶ cells/well in 24-well plates and cultured in RPMI 1640 medium (Thermo Fisher Scientific, Cat. No. 11875093) supplemented with 10% FBS (Thermo Fisher Scientific, Cat. No. A5256701) and 1% penicillin-streptomycin at 37 °C and 5% CO₂ until sub-confluent. Lactiplantibacillus strains were grown overnight in MRS broth (10⁸ CFU/mL), diluted 1:100 (v/v) in serum-free RPMI to ~ 10⁶ CFU, centrifuged at 4000 × g for 3 min, and the supernatants were used for treatment. Controls included MRS medium (vehicle), untreated cells (negative), and 1% Triton X-100 (positive). After 24 h, LDH release was quantified to assess membrane damage (Zipperer et al. 2023). Supernatants (50 µL) were mixed with LDH reaction mix, incubated for 30 min in the dark, and absorbance was measured at 490 nm (reference 620 nm). Cytotoxicity (%) was calculated relative to Triton X-100-treated cells. Parallelly, cell viability was determined using the MTT assay. After treatment removal, 10 µL MTT reagent was added, incubated for 4 h, followed by solubilization and overnight incubation. Absorbance was read at 570 nm (reference 630–690 nm), and viability (%) was calculated relative to untreated controls. All experiments were performed using three independent biological replicates, each measured in technical triplicate. Statistical analysis was conducted using one-way ANOVA with Tukey post hoc test (p < 0.05). This dual approach enabled a comprehensive evaluation of mitochondrial activity and membrane integrity, supporting the safety assessment of Lactobacillus strains on intestinal cells.

Immunomodulatory assay

To evaluate the immunomodulatory effects of Lactiplantibacillus culture supernatants on host immune responses, we measured the secretion of interleukin-1 beta (IL-1β) and interleukin-10 (IL-10) in human epithelial cells using enzyme-linked immunosorbent assay (ELISA) kits from Invitrogen. This protocol was designed to assess both pro-inflammatory and anti-inflammatory responses triggered by bacterial strains and metabolites in vitro (De Marco et al. 2018). The strains were grown overnight in MRS broth at 37 °C under anaerobic conditions until reaching approximately 10⁸ CFU/mL. Cultures were diluted 1:100 in antibiotic-free RPMI. Caco-2 (HTB-37 ™, ATCC) cells were seeded in 24-well plates at a density of 1 × 106 cells per well and cultured in complete medium without antibiotics. The cells were then treated with 1/100 (v/v) dilutions of Lactiplantibacillus strains for 24 h. MRS medium alone was used as a vehicle control. As a positive control for cytokine induction, Escherichia coli (E. coli ATCC11229, final concentration ~ 10⁷ CFU/mL) was used to stimulate a robust innate immune pro-inflammatory response, particularly IL-1β secretion. This bacterial control was included to benchmark the immunostimulatory potential of Lactiplantibacillus supernatants against a known pro-inflammatory stimulus. After incubation, culture supernatants were harvested and centrifuged at 300 × g for 5 min to remove cellular debris, then stored at −20 °C until ELISA analysis. Quantification of IL-1β and IL-10 was performed using human-specific ELISA kits (Catalog # KHC0011 and EHIL10, Invitrogen, Waltham, MA), following the manufacturer’s instructions. Briefly, standards and cell supernatants were added to pre-coated 96-well plates and incubated at room temperature or overnight at 4 °C. After washing, detection antibodies were added, followed by incubation with enzyme conjugates. Colorimetric detection was performed using tetramethylbenzidine (TMB) substrate, and the reaction was stopped with sulfuric acid. Absorbance was measured at 450 nm using a microplate reader. Cytokine concentrations were quantified directly from culture supernatants collected per well and were calculated based on standard curves generated using known concentrations of recombinant IL-1β and IL-10. Values were reported as pg/mL. All experiments were carried out in triplicate and repeated three times. Statistical analyses were performed using one-way ANOVA followed by appropriate post hoc testing to determine significant differences between groups.

Effect of external metabolites toward Staphylococcus aureus ATCC1026 using transmission and scanning electron microscopy (TEM and SEM)

The ultrastructural effects of LAB external metabolites on S. aureus ATCC1026 were examined using both TEM and SEM. Exponentially growing bacterial cells (1 × 10⁶ CFU/mL) were independently treated with 1 × MIC concentrations of Met-ext obtained from each strain for 6 h at 37 °C, following previously described protocols (Tenea 2020). For TEM analysis, treated cells were fixed and embedded, and ultrathin sections were prepared, mounted on copper grids, and stained with uranyl acetate and lead citrate (Sigma-Aldrich, St. Louis, MO, USA). Ten random sections per treatment were examined using a Tecnai G2 F20 transmission electron microscope (FEI Company, Hillsboro, OR, USA) to assess intracellular alterations. For SEM analysis, both treated and untreated cells were resuspended in PBS, air-dried at room temperature, and fixed with 2.5% glutaraldehyde overnight at 4 °C. Samples were washed thrice with phosphate buffer and once with distilled water, then dehydrated with graded ethanol concentrations (50%, 75%, 95%, and 100%, 15 min each). After critical point drying (CPD), samples were mounted on graphite tape and sputter-coated with a ~ 24.5 nm layer of gold using a DENTON VACUUM Desk IV coater (DENTON VACUUM, Austin, TX, USA). Morphological and topographical features were visualized using a JSM-6490 LV scanning electron microscope (JEOL, MA, USA) under high vacuum conditions with secondary electron detection at an accelerating voltage of 15 kV. For semi-quantitative consistency, a minimum of 10 randomly selected fields per sample and at least 100 individual cells per treatment were evaluated.

Results

Strain-specific internal metabolites underpin probiotic and antimicrobial functions

LC–MS/MS analysis identified a total of 117 Met-Int across the three L. plantarum strains, demonstrating both conserved and strain-specific biochemical features relevant to probiotic function and antimicrobial activity (Tables 1, 2, and 3). Common metabolites such as l-tryptophan, detected in all three strains, as well as melibiose (Gt2 and Gt3), adenosine (Gt3 and Gt28L), and NAD⁺ (Gt2 and Gt28L), suggest the presence of conserved pathways involved in host immune regulation, energy metabolism, and microbial modulation. Beyond these shared features, each strain demonstrated distinctive internal metabolomic characteristics reflective of their multifunctionality. Gt2 stood out for its diverse and bioactive profile, including compounds like tanshinone IIA, crassostreaxanthin A, tetramethylscutellarein, and 2-acetoxy-4-pentadecylbenzoic acid (Table 1), which are associated with antioxidant, anti-inflammatory, and antimicrobial activities. Gt3 exhibited enrichment in metabolites such as tyrosine, fructosyl isoleucine, and daidzein (Table 2), supporting roles in neurotransmitter biosynthesis, metabolic adaptability, and immunomodulation. Meanwhile, Gt28L was characterized by potent antimicrobial and signaling molecules, including gancidin W, 1-indole-3-carboxaldehyde, indoleacrylic acid, indole, and cyclic dipeptides (Table 3), suggesting mechanisms of inter-bacterial communication and host immune modulation. Together, these findings underscore the metabolic diversity and functional specialization of L. plantarum strains, informing their potential applications in gut health and antimicrobial therapy.

Table 1.

Metabolites from Gt2 with antimicrobial and probiotic activity

Compound	Antimicrobial and probiotic activity	Notes
Tanshinone IIA	Gut health modulation + strong antimicrobial activity	Found in Salvia miltiorrhiza; modulates gut microbiota and inhibits bacteria/fungi (Fang et al. 2021)
Melibiose	Prebiotic (promotes beneficial microbiota growth) + mild antimicrobial	Acts as a prebiotic, enhancing intestinal beneficial bacteria; mild antimicrobial effects (İspirli et al. 2020)
beta-Nicotinamide adenine dinucleotide (NAD+)	Supports gut health (metabolism); no direct antimicrobial	Key coenzyme in metabolism, supports probiotic growth via redox balance and immunity (Xu et al. 2020)
l-Tryptophan	Probiotic (AhR pathway activation) + antimicrobial (indole derivatives)	Boosts gut immunity and microbiota through AhR and its indole derivatives, which have antimicrobial activity (Roager and Licht 2018)
Tetramethylscutellarein	Antioxidant, gut health support + antimicrobial	A flavonoid with antimicrobial and gut-modulating effects (Wang et al. 2022)
8-(2,3-Dihydroxy-3-methylbutyl)−7-methoxychromen-2-one	Emerging prebiotic + antimicrobial (flavonoid)	Flavonoid-like, supports gut health and inhibits pathogens (Fischer 2020)
Crassostreaxanthin A	Possible gut health modulation + antimicrobial	Carotenoid with antimicrobial and anti-inflammatory properties; potential probiotic effect under investigation (Sinha et al. 2023)
Daidzein	Gut microbiome modulator + mild antimicrobial	Isoflavonoid improving microbiota composition and weakly inhibiting pathogens (Atkinson et al. 2005)
Indaconitine	Non-probiotic + strong antimicrobial	Potent antimicrobial alkaloid but toxic, mainly pharmacological interest (Yu et al. 2021)
2-Acetoxy-4-pentadecylbenzoic acid	Gut health modulation + antimicrobial (fatty acid derivative)	Fatty acid derivative showing antimicrobial activity and potential gut microbiome modulation (Casillas-Vargas et al. 2021)

Table 2.

Metabolites from Gt3 with antimicrobial and probiotic activity

Compound	Antimicrobial and probiotic activity	Notes
Tyrosine	Precursor for tyramine; minimal direct antimicrobial effect	Fermented by probiotics; tyramine is a biogenic amine with neuromodulator roles, but minimal direct pathogen kills (Bonnin-Jusserand et al. 2012)
Methionine	Supports gut barrier and redox balance; minimal antimicrobial	Crucial for mucosal immunity and antioxidative defense (Kachungwa Lugata et al. 2022), but lacks direct antimicrobial action
l-Phenylalanine	Microbial metabolism precursor; minimal direct antimicrobial	Metabolized to phenylacetic acids with mild antimicrobial properties (Qiu et al. 2024)
l-Tryptophan (L-Trp)	Activates AhR pathway; boosts gut immunity; indirect antimicrobial	Indole derivatives (e.g., indole-3-aldehyde) enhance barrier defense and inhibit pathogens (Wang et al. 2023)
N-Fructosyl isoleucine	Emerging probiotic metabolite; no clear antimicrobial role	Found in gut environments; supports beneficial microbial metabolism (current emerging research)
Daidzein	Modulates gut microbiota; mild antimicrobial	Isoflavonoid improving microbiota composition and weakly inhibiting pathogens (Atkinson et al. 2005)

Table 3.

Metabolites from Gt28L with antimicrobial and probiotic activity

Compound	Antimicrobial and probiotic activity	Notes
1-Indole-3-carboxaldehyde	Probiotic trait	Activates AhR; enhances gut barrier and immunity (Zelante et al. 2013)
Indole	Antimicrobial + probiotic trait	Quorum sensing modulator; gut barrier protector (Bansal et al. 2010)
Indoleacrylic acid	Antimicrobial + probiotic trait	Anti-inflammatory, protects against pathogens (Wlodarska et al. 2017)
4-Hydroxybenzaldehyde	Antimicrobial	Phenolic antimicrobial agent (Lou et al. 2012)
Cyclo(leucylprolyl)	Antimicrobial	Diketopiperazine; inhibits pathogens (Yan et al. 2004)
Cyclo(D-Arg-L-Pro)	Antimicrobial	Cyclic dipeptide with broad-spectrum activity (Holden et al. 1999)
Maculosin	Antimicrobial	Antimicrobial diketopiperazine (Shaala et al. 2019)
Gancidin W	Antimicrobial	Known antibiotic from Streptomyces (Thuy et al. 2010)
Tyramine	Mild antimicrobial	Biogenic amine; produced by probiotics (Guryanova 2023)
3-Oxo-tetradecanoyl-homoserine lactone	Quorum sensing signal (target for disruption)	Targeted by probiotics to inhibit pathogen QS (Chung et al. 2011)
9,12,15-Hexadecatrienoic acid	Antimicrobial	Fatty acid; membrane-disrupting activity (Desbois and Smith 2010)
7-Octadecenoic acid	Antimicrobial	Unsaturated FA with antimicrobial effect (Desbois and Smith 2010)
2,4,8-Eicosatrienoic acid	Antimicrobial	Polyunsaturated FA; disrupts bacterial membranes (Wang et al. 2020)
8,11-Eicosadienoic acid	Antimicrobial	Similar FA antimicrobial mechanism (Desbois and Smith 2010)
2-Hydroxycinnamic acid	Antimicrobial	Phenolic acid with antibacterial activity (Lou et al. 2012)
Cyclo(-His-Pro)	Antimicrobial	Quorum sensing inhibitor and antimicrobial agent (Otvos 2005)
Cyclo(-Pro-Val)	Antimicrobial	Reported in antibacterial activity studies (Corrêa et al. 2023)
Prolyl-Valine, Tyrosyl-Valine, Prolyl-Tyrosine	Possible activity	Small dipeptides: some reported antimicrobial roles (Corrêa et al. 2023)
Leucyl-Leucine	Possible activity	Dipeptide with weak antimicrobial reports (Corrêa et al. 2023)

Extracellular metabolomic profiling reveals diverse and strain-specific antimicrobial signatures

Untargeted LC–MS/MS profiling of the Met-Ext from L. plantarum Gt2, Gt3, and Gt28L identified 32 extracellular metabolites, encompassing both shared and strain-specific compounds with putative antimicrobial activity. Presence/absence clustering revealed clear metabolite partitioning among strains, with certain Met-Ext detected across multiple isolates and others uniquely associated with a single strain (Fig. 1).

Fig. 1.

Comparative heatmap of unique metabolite profiles detected in L. plantarum strains Gt2, Gt3, and Gt28L by untargeted LC–MS analysis. Each row corresponds to an individual metabolite, with red cells indicating its presence and blue cells indicating its absence in the respective strain

Notably, l-tryptophan was found in all strains, while l-methionine, l-tyrosine, pseudo-anisatin, and loperamide were shared between Gt3 and Gt28L. Surfactant-like molecules such as N,N-dimethyldodecylamine N-oxide and N,N-dimethyltetradecylamine, along with the cinnamic acid derivative 1-(3,4-dimethoxycinnamoyl)piperidine, were also common to these two strains, suggesting overlapping antimicrobial potential. Gt2 uniquely produced lincomycin, quinidine, methoxyindoleacetic acid, and a complex glycoside, indicating a targeted antimicrobial profile. Gt3 displayed the broadest spectrum of antimicrobial metabolites, including alkaloids (harmine, norharmane), flavonoids (genistein, daidzein), and additional bactericidal agents such as cinnamic acid derivatives and dodemorph. In contrast, Gt28L harbored distinctive antimicrobial compounds such as baicalein, 2-isopropylthioxanthone, 4-(dimethylamino)pyridine, and a chromen-4-one derivative, highlighting a unique yet potent bioactive signature. Metabolite enrichment analysis further emphasized these differences in functional capacity. Gt2 exhibited modest enrichment in nicotinate and nicotinamide metabolism, amino acid biosynthesis (glycine/serine/threonine; phenylalanine/tyrosine/tryptophan), and galactose metabolism, although none reached significance after correction (Fig. 2A). In contrast, Gt3 showed a stronger signal, with phenylalanine, tyrosine, and tryptophan biosynthesis emerging as the most significantly enriched route (adjusted p = 0.0167), accompanied by contributions from novobiocin biosynthesis, phenylalanine metabolism, and additional amino acid pathways (Fig. 2B). Similarly, Gt28L displayed enrichment in nicotinate and nicotinamide metabolism, phenylalanine, tyrosine, and tryptophan biosynthesis, phenylalanine metabolism, and novobiocin biosynthesis, alongside signals in purine and one-carbon metabolism (Fig. 2C). A detailed summary of the pathway enrichment analysis, including the identified metabolic routes, the number of metabolites detected within each pathway, and their associated statistical parameters, is provided in Tables S1–S3. Altogether, these results suggest that aromatic amino acid metabolism represents a conserved and biologically relevant route among the Amazonian strains, potentially linked to the production of bioactive metabolites with antimicrobial or antioxidant activity, while Gt3 and Gt28L may harbor unique metabolic capabilities compared to Gt2.

Fig. 2.

Metabolite set enrichment overview. Bubble size represents pathway enrichment, and color indicates pathway classification. A Gt2 Met-Int; B Gt3 Met-Int; C Gt28L Met-Int. The x-axis denotes the enrichment ratio, reflecting the degree of overrepresentation relative to background levels, while bubble coloration corresponds to statistical significance (−log10 p), with red indicating the highest significance and yellow the lowest. Key pathways affected include nicotinate and nicotinamide metabolism, phenylalanine, tyrosine and tryptophan biosynthesis, and glycine, serine, and threonine metabolism, among others

Lactiplantibacillus strains exhibit high biocompatibility and preserve epithelial cell integrity

The cytotoxicity assessment of extracellular metabolites from L. plantarum strains Gt2, Gt3, and Gt28L revealed excellent biocompatibility with human colon epithelial cells, as demonstrated by MTT and LDH assays. The MTT assay (Fig. 3) indicated that all three strains maintained cell viability above 85%, with Gt28L and Gt2 achieving levels comparable to the untreated control (~ 100%). Gt3-treated cells showed slightly lower viability but remained within the acceptable, non-cytotoxic range. In contrast, MRS medium reduced cell viability to approximately 60–70%, and Triton X-100 (positive control) caused near-complete loss of viability, confirming assay sensitivity.

Fig. 3.

Comparative analysis of cell viability following treatment with various LAB strains (1/100, v/v dilution). Cell viability was assessed using the MTT assay and is expressed as a percentage relative to untreated control cells. The solid light-gray bar represents the negative control (untreated cells), while the white bar corresponds to the positive cytotoxic control (1% Triton X-100). Patterned bars (dotted, diagonal-striped, and fine-dotted) represent LAB strains exhibiting moderate to good biocompatibility, whereas the horizontally striped bar indicates a strain associated with reduced cell viability. Data are shown as mean ± SEM from three independent experiments. Asterisks indicate statistically significant differences compared to the cytotoxic control (***p < 0.001, one-way ANOVA with Tukey’s post hoc test)

Similarly, LDH assay results (Fig. 4) supported the non-cytotoxic nature of the strains. All LAB treatments led to LDH release below 10%, akin to the negative control, while the MRS medium caused a modest increase in LDH but still within a non-toxic range. In comparison, the positive control Triton X-100 induced nearly 100% LDH release, with statistically significant differences (p < 0.001) from all other groups. Collectively, these findings confirm that postbiotic metabolites from Gt2, Gt3, and Gt28L preserve both cellular viability and membrane integrity, validating their safety for potential probiotic and postbiotic applications.

Fig. 4.

LDH cytotoxicity assay evaluating the effects of LAB strains on Caco-2 cell membrane integrity. LDH release is expressed as a percentage relative to the positive cytotoxic control. The solid white bar represents the positive control (1% Triton X-100), corresponding to maximum LDH release. The solid light-gray bar denotes the negative control (untreated cells). The remaining patterned bars (dotted, diagonal-striped, fine-dotted, and horizontally striped) correspond to individual LAB strains. Data are shown as mean ± SEM from three independent experiments. Asterisks indicate statistically significant differences compared to the cytotoxic control (***p < 0.001, one-way ANOVA with Tukey’s post hoc test)

Strain-dependent immunomodulatory effects of Lactiplantibacillus metabolites on cytokine secretion

Cytokine analysis demonstrated distinct immunomodulatory profiles elicited by the Lactiplantibacillus strains, highlighting their strain-specific effects on inflammatory responses. For the pro-inflammatory cytokine IL-1β (Fig. 5), Gt2 and Gt3 induced secretion levels of approximately 30 pg/mL, comparable to MRS (≈30 pg/mL), corresponding to a threefold increase relative to untreated control cells (≈10 pg/mL) (p < 0.05, n = 3 biological replicates × 3 technical replicates). In contrast, Gt28L reduced IL-1β secretion to ~ 15 pg/mL, representing a 50% decrease compared with MRS (p < 0.05) and a ~ 1.5-fold increase relative to the untreated control. As expected, stimulation with E. coli increased IL-1β to ~ 40 pg/mL, a fourfold induction compared with untreated cells (p < 0.05), confirming assay responsiveness.

Fig. 5.

Induction of IL-1β (pg/mL) secretion by L. plantarum strains Gt2, Gt3, and Gt28L in Caco 2-stimulated cells. MRS medium and untreated cells (control) served as baseline references, while E. coli ATCC11229 was used as a pro-inflammatory positive control. Data are shown as mean ± SEM from three independent experiments. Asterisks indicate statistically significant differences between the untreated control and the E. coli–stimulated cells (*p < 0.05, one-way ANOVA with Tukey’s post hoc test)

Conversely, the anti-inflammatory cytokine IL-10 (Fig. 6) was most strongly induced by Gt28L, which increased IL-10 secretion to approximately 260–270 pg/mL, corresponding to a ~ 6.5-fold increase relative to the untreated control (~ 40 pg/mL) (p < 0.01). Gt2 and Gt3 also elevated IL-10 production to ~ 170–180 pg/mL, representing ~ 4.3–4.5-fold increases compared with control cells (p < 0.05). In contrast, MRS-treated cells showed only ~ 50 pg/mL and untreated controls ~ 40 pg/mL, while E. coli stimulation resulted in minimal IL-10 secretion (< 30 pg/mL), consistent with its pro-inflammatory profile. These results underscore the capacity of Lactiplantibacillus metabolites, particularly from Gt28L, to modulate immune responses by attenuating pro-inflammatory signals while enhancing anti-inflammatory pathways.

Fig. 6.

IL-10 (pg/mL) secretion following stimulation with L. plantarum strains Gt2, Gt3, Gt28L, and E. coli. MRS medium and untreated cells (control) were included as controls, while heat-killed E. coli ATCC11229 was used as a pro-inflammatory reference. Data are shown as mean ± SEM from three independent experiments. Gt28L showed a statistically significant increase in IL-10 secretion compared to all other groups (**p = 0.0007). Strain-specific comparisons also showed significant differences for Gt2 vs. control (*p = 0.0147) and Gt3 vs. control (*p = 0.0142)

Ultrastructural damage to S. aureusinduced by Lactiplantibacillus extracellular metabolites

SEM and TEM analyses demonstrated significant morphological disruptions in S. aureus ATCC1026 cells treated with extracellular metabolites (Met-Ext) from the Lactiplantibacillus strains Gt2, Gt3, and Gt28L (Fig. 7). Untreated control cells maintained a typical coccoid shape with smooth, intact surfaces and dense cytoplasm. In contrast, all Met-Ext treatments induced pronounced structural damage: Gt2 caused cell elongation, surface collapse, and severe lysis, resulting in ghost-like cell remnants; Gt3 exposure led to membrane rupture and cytoplasmic vacuolization; and Gt28L-treated cells exhibited swelling and electron-lucent cytoplasm, indicative of disrupted membrane integrity and impaired intracellular metabolism. These observations confirm a strain-dependent but convergent antibacterial mechanism primarily targeting the bacterial cell envelope and internal architecture through membrane disruption.

Fig. 7.

Ultrastructural effects of extracellular metabolites (Met-Ext) from Gt2, Gt3, and Gt28L strains on S. aureus cells observed by SEM (top panel) and TEM (bottom panel). Untreated control cells show intact spherical morphology and dense cytoplasm. Met-Ext treatments from Gt2, Gt3, and Gt28L strains caused membrane disruption, surface irregularities, cytoplasmic leakage, and structural collapse. Besides, Gt2 treatment resulted in pronounced lysis and formation of ghost-like cells. Scale bars: 1 µm (SEM) and 500 nm (TEM)

Discussion

Distinct internal metabolomic signatures underpin probiotic and antimicrobial potential of L. plantarumstrains

The intracellular metabolic profiles of the three bacterial strains reveal distinct compounds that may underpin their probiotic and antimicrobial activities, providing insights into each strain’s capacity to support gut health, modulate immune responses, and inhibit pathogenic microorganisms. Several metabolites were common across the strains, highlighting potential shared mechanisms for probiotic activity (Tables 1, 2, and 3). Among the identified compounds, melibiose, l-tryptophan, and adenosine stand out for their direct relevance to probiotic activity. Melibiose, a prebiotic disaccharide detected in Gt2 and Gt3, is known to stimulate the proliferation of beneficial gut microbiota such as Lactobacillus and Bifidobacterium, thereby promoting microbial balance and enhancing gut barrier integrity (Gänzle and Follador 2012). l-Tryptophan, an essential amino acid, serves as a precursor for bioactive indole derivatives that regulate mucosal immunity, reinforce epithelial defense, and control intestinal inflammation via activation of aryl hydrocarbon receptors (Zelante et al. 2013). Adenosine, detected in Gt3 and Gt28L, further complements these effects through its established role in dampening excessive inflammatory responses in the gut, supporting homeostasis and protecting against conditions like inflammatory bowel disease (Antonioli et al. 2013). Its presence suggests that these strains may have a role in regulating energy homeostasis and immune modulation, which is essential for maintaining gut integrity and preventing inflammation. Moreover, the presence of NAD+ in Gt2 and Gt28L indicates that these strains may have enhanced resilience to environmental stressors, such as changes in pH or nutrient availability, which could improve their survival and efficacy as probiotics (Covarrubias et al. 2021). The combination of these metabolites suggests that Gt3 may provide beneficial effects on the gut microbiota while supporting host health through immune modulation and metabolic balance. Together, these metabolites suggest that Gt2, Gt3, and Gt28L may possess probiotic traits that help promote a healthy gut microbiota, enhance the immune response, and potentially improve overall gut function. In addition, the Gt2 strain exhibits a diverse and robust metabolite profile that underscores its potential as a multifunctional probiotic (Table 1). Recent findings by Molina et al. (2025) provide the first comprehensive metabolomic characterization of Gt2, confirming the presence of several bioactive compounds with protective effects on gut health and notable antimicrobial potential. Among its key metabolites, tanshinone IIA, a lipophilic diterpene, exhibits strong antimicrobial and antioxidant activities, which may help mitigate oxidative stress and support mucosal defense mechanisms (Lili et al. 2023). The presence of crassostreaxanthin A, a carotenoid known for its antioxidant capacity (Maoka 2020), and tetramethylscutellarein, a flavonoid with anti-inflammatory effects (Liu et al. 2025), further enhances the strain’s ability to preserve intestinal barrier function and promote microbial homeostasis. While Gt2 demonstrates a balanced profile with host-protective and anti-inflammatory properties, a contrast emerges when comparing it to the metabolomic profiles of Gt3 and Gt28L, which highlight more specialized or enhanced antimicrobial traits (Tables 2 and 3). Gt3 is distinguished by its enrichment in metabolites such as tyrosine, a precursor of neurotransmitters like dopamine, suggesting a possible gut-brain axis influence (Chen et al. 2021). The detection of fructosyl isoleucine implies heightened metabolic adaptability and stress resistance, both important for probiotic persistence. Additionally, the presence of daidzein, an isoflavonoid with antimicrobial and immunomodulatory properties (Atkinson et al. 2005), supports Gt3’s ability to modulate gut microbiota, a role reinforced by Xu et al. (2025), who linked similar metabolites to beneficial shifts in microbial composition. In comparison, Gt28L expresses a more complex profile, integrating antimicrobial compounds like gancidin W, effective against Gram-positive bacteria (Thuy et al. 2010), and indole derivatives such as 1-indole-3-carboxaldehyde. These molecules play crucial roles in inter-bacterial communication and host immune modulation (Zelante et al. 2013). The unique combination of immune-regulating and bactericidal metabolites in Gt28L supports its dual functionality, simultaneously promoting host health and suppressing pathogens. Thus, these strain-specific metabolomic signatures highlight the functional diversity within L. plantarum isolates. Their distinct combinations of metabolites underline their potential as targeted probiotics capable of not only sustaining gastrointestinal health but also offering natural antimicrobial solutions.

Strain-specific extracellular metabolites reveal broad-spectrum and targeted antimicrobial capacities in L. plantarum isolates

External metabolite profiling revealed both shared and unique compounds across the three L. plantarum strains, highlighting their distinct antimicrobial capacities. l-Tryptophan, consistently detected across all strains, has been implicated in maintaining intestinal barrier integrity and modulating host immune responses (Wu 2009), while its indole derivatives play a pivotal role in shaping microbial community dynamics (Wikoff et al. 2009). l-Methionine and l-tyrosine, detected in Gt3 and Gt28L, are essential amino acids involved in fundamental cellular pathways and may act as precursors for the biosynthesis of antimicrobial or signaling molecules. Moreover, the shared presence of pseudo-anisatin and loperamide, both reported to exert moderate antimicrobial effects (Butler 2008), suggests a baseline functional overlap between Gt3 and Gt28L (Fig. 1). The identification of membrane-disrupting agents like N,N-dimethyldodecylamine N-oxide and 1-(3,4-dimethoxycinnamoyl)piperidine further indicates potential for broad-spectrum antimicrobial activity (Thompson et al. 2022; Alshibl et al. 2020). Gt2 exhibited significant biosynthetic versatility, producing lincomycin, a canonical lincosamide antibiotic (Spízek and Rezanka 2004), alongside quinidine, a compound with potent antifungal and antiparasitic activity (Upadhyay et al. 2014). The detection of methoxyindoleacetic acid aligns with indole-mediated antimicrobial mechanisms (Babalola et al. 2025), suggesting that Gt2 metabolites may disrupt quorum sensing and microbial communication networks. Gt3, in contrast, displayed a chemically diverse antimicrobial repertoire, notably enriched in β-carboline alkaloids such as harmine and harmane, which exert broad-spectrum effects through enzyme inhibition and DNA intercalation (Carvalho et al. 2017). The co-occurrence of flavonoids genistein and daidzein further potentiates antimicrobial activity by modulating bacterial quorum sensing and impairing biofilm formation (Verdrengh et al. 2004), collectively highlighting Gt3 as a robust source of multifunctional antimicrobial metabolites. These, along with membrane-disrupting surfactants and cinnamic acid derivatives (Allegrone et al. 2021), position Gt3 as a multifunctional strain with wide-ranging antimicrobial properties. Gt28L, while narrower in metabolite number, carried highly potent compounds. Baicalein, with its antibacterial and antiviral activities (Hu et al. 2022), and 2-isopropylthioxanthone, known for photodynamic antimicrobial action (Al-Azawi et al. 2025), demonstrate the unique antimicrobial profile of this strain. Additionally, the presence of 4-(dimethylamino)pyridine and apigenin derivatives, which inhibit bacterial adhesion and inflammation (Pei et al. 2023), underscores its therapeutic relevance. Metabolite enrichment analysis substantiated the strain-specific antimicrobial profiles by revealing distinct biosynthetic pathway activations (Fig. 2). In Gt2, the prominent engagement of nicotinate and nicotinamide metabolism suggests a metabolic orientation toward maintaining redox balance and modulating host immune responses (LeBlanc et al. 2013; Cervantes-Barragan et al. 2017). Gt3 exhibited significant enrichment of aromatic amino acid biosynthesis and novobiocin biosynthetic pathways, consistent with the synthesis of bioactive antimicrobial compounds, including antibiotics (Grossman et al. 2023; Arena et al. 2016). In contrast, Gt28L demonstrated broad activation across several biosynthetic routes, particularly those involved in folate and one-carbon metabolism, indicating a dual role in supporting host cellular processes and exerting antimicrobial effects (Carboni 2022). Overall, the comparative metabolomic evidence positions Gt3 as the most promising candidate due to its wide array of antimicrobial compounds. Gt28L follows with a more targeted but potent profile, and Gt2 remains valuable for its defined yet effective antimicrobial signature. These results align with previous studies (Tenea and Ortega 2021) and support the use of these strains in antimicrobial applications such as food preservation, gut health modulation, and natural alternatives to antibiotics. Moreover, the diversity of antimicrobial metabolites emphasizes the relevance of secondary metabolites, not just bacteriocins, in defining probiotic efficacy (O’Callaghan and van Sinderen 2016).

Strain-specific biocompatibility and proliferative support of colon epithelial cells by Lactiplantibacillus metabolites

The results from the MTT assay demonstrate that all tested Lactiplantibacillus strains are non-cytotoxic and potentially pro-proliferative, especially Gt28L and Gt2, which supported cell viability levels like untreated controls (Fig. 3). These findings suggest a beneficial interaction with colon epithelial cells. In previous work, L. plantarum isolate Y8 demonstrated significant antiproliferative effects against oral cancer cell lines without exhibiting cytotoxicity toward normal cells, consistent with earlier studies reporting the safety of other L. plantarum strains in vitro (Haghshenas et al. 2023). The high viability may be attributed to bioactive postbiotic compounds such as SCFAs, EPS, and bacteriocins, which have been shown to enhance mitochondrial function and cell metabolism (Ragavan and Hemalatha 2023). Although Gt3 exhibited slightly lower viability, it still fell within the biocompatible range, pointing to strain-specific variations in metabolite production. Interestingly, MRS medium alone negatively impacted cell viability, likely due to its complex composition, which can stress epithelial cells in the absence of beneficial bacterial metabolites. This underscores the necessity of including appropriate controls when assessing postbiotic effects and highlights the potential for bacterial metabolism to detoxify or buffer the medium’s components. The LDH assay further validated the safety of these strains (Fig. 4). The minimal LDH release observed across all bacterial treatments suggests that the strains do not compromise membrane integrity. The MRS medium induced slightly higher LDH release than the bacterial treatments, again indicating a mild background effect. The strong cytotoxic effect of Triton X-100 confirmed the sensitivity of the assay and reinforced the non-cytotoxic profile of the tested strains. Together, the MTT and LDH assays provide a consistent and comprehensive safety profile, confirming that these strains are well tolerated by colon epithelial cells. Their metabolites appear to support not only cellular viability but also membrane stability. These characteristics make Gt2, Gt3, and Gt28L promising candidates for further development as postbiotic ingredients in dietary supplements, functional foods, or therapeutic formulations targeting gut health.

Immunomodulatory and antimicrobial mechanisms of Lactiplantibacillus strains: cytokine regulation and membrane disruption

The cytokine data demonstrate that Lactiplantibacillus strains elicit strain-dependent immunomodulatory effects in vitro. IL-1β secretion served as a marker of pro-inflammatory activity, while IL-10 was used to assess anti-inflammatory potential. Compared to the strong IL-1β induction by E. coli, Gt2 and Gt3 stimulated only moderate IL-1β responses (Fig. 5), while Gt28L significantly reduced IL-1β secretion, pointing to a dampened inflammatory profile (Brauner et al. 2001; Alipour et al. 2025). Conversely, Gt28L exhibited the highest induction of IL-10, a key cytokine in promoting mucosal immune tolerance and regulating inflammation (Fig. 6). Gt2 and Gt3 also elevated IL-10 secretion above the baseline, although to a lesser extent. This suggests that all three strains may contribute to immune homeostasis, with Gt28L being the most potent in activating anti-inflammatory pathways. Minimal IL-10 levels from E. coli and controls further confirmed the immunomodulatory specificity of Lactiplantibacillus metabolites. These findings align with prior studies indicating that postbiotic components, such as indole derivatives or SCFAs, can engage host pattern recognition receptors to modulate cytokine release (Round and Mazmanian 2009; Hu et al. 2015). The reduced IL-1β alongside elevated IL-10 secretion by Gt28L highlights its potential utility in conditions characterized by mucosal inflammation or immune dysregulation. As such, this strain may be particularly suitable for applications targeting inflammatory bowel diseases or maintaining epithelial immune tolerance (Yoo et al. 2020; Tobias et al. 2025). Altogether, the observed cytokine profiles underscore the potential of these L. plantarum strains to fine-tune immune responses. Their ability to modulate both pro- and anti-inflammatory pathways supports their candidacy for further development as next-generation probiotics with immunoregulatory properties. Despite these promising findings, several limitations should be acknowledged. The functional assessments were conducted under in vitro conditions, which may not fully reflect the complexity of in vivo gut environments. The immune-modulatory and cytoprotective effects observed in epithelial cell lines require validation in animal models to confirm physiological relevance.

Strain-specific membrane disruption as a core mechanism of antimicrobial activity

The observed ultrastructural damage to S. aureus cells confirms that extracellular metabolites from L. plantarum strains act via a bactericidal mechanism involving direct disruption of membrane integrity. The SEM and TEM findings, membrane rupture, surface deformation, cytoplasmic leakage, and cell collapse (Fig. 7), are consistent with prior reports that describe similar damage caused by Lactiplantibacillus-derived antimicrobial compounds (Fhoula et al. 2023; Agrawal et al. 2020). These morphological features suggest that Met-Ext fractions compromise cell wall integrity, leading to lysis and functional collapse. Gt2 appeared particularly potent, causing ghost-like cell formation, which may indicate the complete disintegration of cytoplasmic content. This aligns with known actions of plantaricins and other L. plantarum–produced antimicrobial peptides that destabilize membrane potential and compromise cell permeability (Sikandar et al. 2022). Furthermore, our results correspond with the findings of Yang et al. (2021), who demonstrated that L. plantarum metabolites induce oxidative stress and structural disintegration in S. aureus. In another study, extracts from L. paracasei showed inhibitory effects against S. aureus ATCC25933 due to their metabolites and various antimicrobial compounds, including small peptides, bacteriocins, and organic acids, such as butyric, acetic, and lactic acid (Shahverdi et al. 2023). Taken together, these data support the role of L. plantarum–derived postbiotics as effective agents capable of targeting bacterial membranes, a mode of action particularly valuable against antibiotic-resistant pathogens (Shahverdi et al. 2023). The strain-specific differences in severity and mode of structural disruption may reflect varying metabolite compositions, underscoring the need for further characterization of individual antimicrobial compounds. Nevertheless, the consistent pattern of membrane-targeted damage across all strains reinforces the potential of these postbiotic formulations as alternative antimicrobial interventions.

Besides, a key advance in our study is the dual profiling of both Met-Int and Met-Ext metabolite pools across three distinct L. plantarum strains, which provides a more holistic understanding of strain-specific antimicrobial potential. Prior studies typically focused on a single strain or analyzed only secreted metabolites, limiting insights into how intracellular biosynthetic capacity complements extracellular antimicrobial action. By comparing both metabolite pools, we can correlate specific intracellular precursors with secreted bioactive compounds, revealing the metabolic networks that underpin functional diversity and efficacy. The strain-specific differences in the severity and type of membrane disruption likely reflect these unique metabolite compositions, emphasizing that not all L. plantarum isolates confer equivalent antimicrobial effects. However, this approach highlights the multifaceted mode of action of L. plantarum–derived postbiotics and underscores the potential of selecting targeted strains with complementary Met-Int/Met-Ext profiles for applications against antibiotic-resistant pathogens.

Conclusions

This study demonstrates that Gt2, Gt3, and Gt28L strains possess distinct metabolic and functional profiles linked to probiotic, immunomodulatory, and antimicrobial potential. Shared metabolites support immune modulation, gut integrity, and inflammation control, while strain-specific compounds highlight unique strengths: Gt2 is rich in antioxidants and antimicrobials; Gt3 combines a broad antimicrobial spectrum with stress-response metabolites; and Gt28L integrates immunoregulatory and membrane-active compounds. Pathway enrichment highlights a redox-oriented profile for Gt2, broad biosynthetic versatility for Gt3, and dual metabolic functionality for Gt28L. All three strains are non-cytotoxic and support epithelial cell growth, with Gt28L exhibiting the strongest anti-inflammatory potential and Gt3 demonstrating the most extensive antimicrobial activity, collectively positioning these isolates as promising, strain-specific probiotic candidates. Furthermore, ultrastructural studies confirmed that extracellular metabolites compromise S. aureus membrane integrity, suggesting a primary antimicrobial mechanism involving membrane disruption. Taken together, these findings highlight the therapeutic promise of these L. plantarum strains as next-generation probiotics and sources of functional postbiotics, particularly for applications targeting epithelial integrity and mucosal immune regulation. Future development may leverage advanced delivery technologies, such as microencapsulation or hydrogel-based formulations, to enhance stability, targeted release, and efficacy of these probiotics and postbiotics in functional foods or therapeutic products.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

G.N.T. designed the study. G.N.T., G.G.P., G.C.M., R.G.P, and V.C contributed to the formal investigation. G.N.T. contributed to conceptualization, methodology, data curation, supervision, project administration, and funding acquisition. G.N.T. contributed to writing-original draft preparation, writing, review, and editing. G.G.P., G.C.M., and R.G.P contributed to writing, review, and editing. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by Universidad Técnica del Norte under grant number No.9674/2024 awarded to GNT. Additionally, GNT received partial support through the Scientific Visitor Fellowship Grant No. 147/2025 from the Research Institute of the University of Bucharest (ICUB), Romania.

Data availability

The datasets generated and analyzed during the current study are included in this published article and its supplementary information files. Certain datasets form part of a pending patent application and are therefore not publicly available. These data are available from the corresponding author on reasonable request and subject to applicable confidentiality requirements.

Declarations

Ethics approval

Not applicable.

Informed consent

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.